« Comments: Spam and not-Spam | Main | Wait, does this mean that Iraq DID have WMDs? »

Tuesday, April 23, 2013

Hunting Zombies: The War on Drugs and Prison Growth

Before jumping in, I’d like to thank Dan for inviting me back to post on Prawfs. I’ve always enjoyed my experiences here in the past, and I’m looking forward to it all once again. Now on to the post.

Zombie ideas—those ideas that have been invalidated by facts but remain widely accepted because of their political appeal—are a menace everywhere. My goal in this post and posts to come is to do my best to kill one of the most persistent zombie ideas in criminal justice.

The zombie I’m going after here? The common claim that the explosion in the US prison population was primarily (or majorly) caused by the War on Drugs.

It was not. Given how overwhelming the data is on this issue, this zombie’s persistence is as puzzling as it is frustrating.

In fact, I’ll go even further: not only has the War on Drugs not been the primary engine of prison growth, it also has not even been the primary engine of racial disparities in prison populations. 

Not that you would know this by looking at the academic and popular literature. Everyone from Michelle Alexander—whose book James Forman accurately dismantles, and which I will be criticizing as well—to Adam Gopnik in the pages of the New Yorker and Fareed Zakaria in the pages of Time tends to make this argument. I took aim at this argument recently in a review of Ernest Drucker’s book, The Plague of Prisons, in Michigan’s book review issue; I want to develop the argument more fully here, and for a wider audience.

Let me be perfectly clear: I am not saying the War on Drugs has played no role in prison growth whatsoever. Nor am I saying that it is racially neutral, nor that it is sufficiently unimportant that we can ignore it. But I am saying that the following statements are wrong, in direct contradiction of what I hear and read all the time:

  1. The incarceration of low-level drug offenders is the primary (or even a major) engine of prison growth.
  2. The incarceration of drug offenders in general is the primary (or even a major) engine of prison growth.
  3. The incarceration of drug offenders is responsible for much of the racial disparity we see in prison populations.
  4. Drug offenders serve disproportionately long sentences.

My guess is that many, if not most, people believe at least some of these statements. But each one is wrong.I will start by demonstrating why each claim is wrong.  Then I’ll show what theories do have empirical support.

When looking at what does cause prison growth, I will include the War on Drugs: there is at least some nuance to my argument here. But while it is true that the War likely matters, its effects are indirect and thus exceedingly hard to measure. Do drug arrests that do not result in incarcerations themselves nonetheless lead to future incarcerations for more-serious crimes by undermining offenders’ career- and other pro-social opportunities? How many aggravated assaults and murders are due to the War’s effect on illegal drug markets? Do drug convictions that do not result in incarceration nonetheless generate longer prior-history records that increase the risk of incarceration for future non-drug crimes?

These are all important ways in which the War on Drugs can matter, and they can matter both in terms of the number of prisoners and the racial or economic composition of prison populations. But in each of these cases, the magnitude of the War’s impact can be hard to disentangle from other factors, and the causal arrow spins madly like a compass on the island from Lost.

I’ll also raise a broader question: how do you actually kill a zombie idea in general? In many ways, zombie is the wrong term to use: we should think of these ideas as hydras.  Zombies are very hard to kill, but each time you shoot/stab/hit/punch a zombie, it gets a bit weaker—just not fast enough. Ironically, as I will discuss later, each time you attack a bad idea it can perversely grow stronger, like the hydra growing two heads for each one chopped off.

The overemphasis of the War on Drugs is not a minor problem. States are eagerly trying to figure out how to reduce their prison populations. If we don’t know how we ended up with so many people in prison, how can we amend our laws and policies we get the outcomes we want? If we act as if the War on Drugs is the cause of our problems, we aren’t going to solve them.

One last thing I want to make clear before I jump into the substance of stuff. I am not a proponent of the War on Drugs. I favor across-the-board legalization (and taxation) of marijuana, and I found the outcome in Gonzales v Raich (upholding federal enforcement against purely intrastate drug markets) depressing. I’m unsure about the extent to which we should legalize or decriminalize harder drugs, but I certainly believe that we should treat drug addiction as a public health problem, not a criminal one, and that we could learn a lot from countries like Portugal. I think the War on Drugs has been, on net, a waste of time, energy, and resources, and that it has been needlessly destructive for whatever it has accomplished.

In other words, I am not a defender of the War on Drugs, nor even some sort of lesser fellow-traveller. I’m pretty much opposed to it. But that doesn’t change the fact that, whatever the War on Drugs has done, it has not driven up our prison populations by incarcerating drug offenders for long periods of time.

Posted by John Pfaff on April 23, 2013 at 09:21 AM | Permalink


TrackBack URL for this entry:

Listed below are links to weblogs that reference Hunting Zombies: The War on Drugs and Prison Growth:


John, I admire your very important work here, but I think you have to make sure you give adequate attention to the reality that "drug convictions that do not result in incarceration nonetheless generate longer prior-history records that increase" GREATLY THE LENGTH of subsequent incarcerations for other offenses.

In lots and lots of federal ACCA cases and California three-strikes cases, drug crimes often serve as critical first/second strikes for subsequent (often minor) crimes that are not thereafter coded as drug crimes but result in long sentences that are functionally a feature of the Drug War. Relatedly, at least in the federal system, underlying drug crimes are the driver for more investigations and convictions on other charges (whether gun charges or immigration charges or...).

Also, as you look at the jail/prison data, you also need to look at the net-widening impact of the drug war that result from (1) parole/probation revocation based on drug issues, (2) mis-allocation of police/probation resources because of drug war concerns.

I am not eager to keep this zombie alive, but I am eager to highlight the ways in which the disease of this zombie infects prison populations in ways both observable and potentiallu unobservable in basic data.

Posted by: Doug B. | Apr 23, 2013 5:17:40 PM

Question for Doug and John or others: to what extent do we think or know that drug prosecutions are in fact prosecutorial low-hanging fruit, that is, charge selection favors even low-level drug charges because it's easier to prove and punish severely some drug crimes even though those who are actually charged and punished for those drug crimes are in fact suspected by police and prosecutors to have engaged in far more severe crimes that are harder to prove.

For example, say Doug B is a cop and JP is a perp :-) Doug thinks JP is involved in a rash of gun violence associated with protecting drug turf, but can't easily prove that whereas through a clever buy and bust, Doug can toss JP into prison for a good few years. I imagine we can't find very good data on this scenario even though I have to think it happens with some frequency, no?

To be sure, this kind of proxy charging rubs me the wrong way, but I'm an institutional retributivist sort of guy who likes the gov't to prove its real case (highest readily provable charge in most situations). But if you're a moral retributivist or just focused purely on the crime control stuff, you might be quite open to this gambit by cops and prosecutors. And if you don't think this practice of proxy charging is so bad, then maybe even the level of incarceration caused by the drug war doesn't bother you so much if you think proxy charging is prevalent.


Posted by: Dan Markel | Apr 23, 2013 10:29:40 PM

@DB: I definitely plan to address the impact of these priors in future posts (and work), and I think your concern is right: this effect is *really* hard to uncover in the data, since we rarely have granular prior-record data. But there is some, so it may be possible to tease out at least some broad conclusions.

The one question I have in return is whether your "for examples" are really *examples.* In other words, the feds and California (at least until Plata) are/were uniquely punitive systems. As Zimring, Hawkins, and Kamin point out, despite the fact that half of all states have a three-strike law, over 90% of all strike inmates are in California. So are Cali and the feds indicative examples, or are the unique outliers?

I have this same problem whenever I read about "the power of prison guard unions, for example in California...." It's always California. It's not a "for example": California's prison guard union is uniquely powerful, and it is dangerous to extrapolate much from it. No idea yet if that is true for drug-prior enhancements for non-drug offenses, so genuinely curious.

@DM: I've often wondered this myself. I think of it less in the drug context, though, and more in the parole revocation setting. People like to say "if we simply stopped revoking parole...." As I'll show later, there are some conceptual problems with this argument in general, but another unraised issue is how many revocations are just shortcuts to avoid a trial for a new substantive offense? No idea. For the drug-non-drug case, you may be able to see something if the non-drug charge is dropped after arraignment (there are some datasets that track from the time of filing), but it is likely hard to uncover as well, at least directly.

Posted by: John Pfaff | Apr 24, 2013 9:33:10 AM

Couple quick follow-ups:

1. DM: My sense is that, in the federal system at least, the low hanging fruit is felon in possession (FIP). And, for repeat offenders, an FIP charge can trigger ACCA punishments of a mandatory minimum of 15 years.

2. JP: I agree 100% that the feds and California are outliers, but significant ones given that these two jurisdictions have, roughly speaking, 25% of the nation's incarcerated population. Indeed, throw in Texas and Florida and we already have nearly 1/2 of the mass incarceration population in the US.

Putting both comments together, this discussion confirms my sense we might really need to dig down jurisdiction-by-jurisdiction to figure this out. Michael O'Hear has a great recent state-specific piece on the data that itself highlights how different the data story can be even in seemingly similar states.

Posted by: Doug B. | Apr 24, 2013 10:15:59 AM

Post a comment